Workers whose schedules lurch from one week to the next may be paying a hidden health cost for that instability, according to a large U.S. study that reveals sharply different consequences for men and women. Researchers at the University of Illinois Urbana-Champaign analyzed employment and health records for more than 61,000 adults and found that volatile work hours were associated with worsening self-rated health among men, while women were more likely to respond to health strain by cutting back their hours, taking time away from work or leaving employment altogether. The findings suggest that unpredictable scheduling is not simply an inconvenience for workers trying to organize family life, transportation or finances. It may also operate as a population-level occupational health risk, with its effects partly concealed when studies examine only people who remain continuously employed.
The study, led by doctoral researcher Sinyee Qianyi Lu and sociology professor Tim Liao, drew on the Current Population Survey, a monthly survey administered jointly by the U.S. Census Bureau and the Bureau of Labor Statistics. The researchers linked the survey’s basic monthly files with annual socioeconomic data from the Integrated Public Use Microdata Series. This allowed them to follow individuals’ reported working hours and health assessments across time rather than taking a single snapshot of employment. Participants were adults between 18 and 64, and the linked records covered a 16-month framework designed to capture month-to-month changes in both work schedules and health. In total, the analysis included workers observed across periods spanning the years before, during and after the COVID-19 pandemic.
A central feature of the research was its focus on work-hours volatility rather than simply the number of hours people worked. A person who works 40 hours every week and another who consistently works 25 hours may have very different schedules, but neither necessarily experiences volatility. By contrast, a worker whose schedule repeatedly jumps from a low number of hours to a high number and then falls again may face substantial instability even if the average number of hours is moderate. The researchers measured these proportional fluctuations, capturing the degree to which an individual’s hours changed between months. This distinction is important because the body may be capable of adapting to a stable workload, while repeated and unpredictable changes can disrupt sleep, recovery, household routines and the ability to plan daily life.
The health measure came from participants’ self-assessments, which ranged from excellent to poor. Rather than treating people who stopped working as missing data, the researchers classified changes into four possible outcomes: improved health, stable health, deteriorated health or a health-related work limitation. The final category included people who began working part time because of illness or medical problems, were absent from work for health reasons or left the labor force because of illness or disability. This approach addressed a major weakness in earlier research. Studies restricted to continuously employed workers can unintentionally remove the people whose health has been affected most severely, leaving behind a relatively healthy group of “survivors” and making the apparent consequences of unstable work look smaller than they are.
In the baseline analysis, 55,062 people remained continuously employed throughout the observation period. An expanded model added 6,069 participants who experienced health-related limitations during the second year. Among men, greater volatility in work hours was linked to a gradual decline in self-rated health, particularly among those working longer weekly hours. The model estimated that the probability of health deterioration rose from about 28% among men whose working hours did not change to approximately 38% among those experiencing extremely large fluctuations between months. Over the same range, the estimated probability of stable health fell from roughly 46% to 38%. The likelihood of reporting improved health changed little, suggesting that volatility was associated primarily with movement away from stability and toward poorer health rather than with a broad reshuffling of all health outcomes.
The pattern among women was less visible when the analysis included only people who stayed employed. Among continuously employed women, work-hours volatility was not clearly associated with either worsening or improving self-rated health. That apparent absence of an effect changed when health-related employment changes were counted as outcomes. Women experiencing health strain were more likely than men to reduce their hours, interrupt employment or leave paid work entirely. As a result, the health consequences of unstable schedules may appear to be weaker for women in conventional analyses simply because some affected workers disappear from the employed sample. The researchers argue that this divergence reflects different pathways through which occupational stress becomes visible: men may remain on the job while reporting deteriorating health, whereas women may be more likely to alter their participation in employment.
The findings do not establish that fluctuating schedules directly caused illness, because the study was based on observational survey data rather than a randomized experiment. Workers with unstable hours may differ from other workers in many ways, including occupation, income, job security, caregiving responsibilities, access to health care or preexisting health conditions. Even so, the longitudinal design strengthens the analysis by examining changes in work hours and subsequent changes in health across multiple observations. The results also remained meaningful after the researchers considered the absolute number of hours worked. According to Liao, the key problem was not necessarily whether a person worked 25 or 40 hours per week, but whether the schedule repeatedly moved up and down. Sudden changes may make it difficult for workers to recover physically, anticipate income or coordinate the demands of work and home.
The gender differences may reflect the unequal structure of paid and unpaid labor as well as differences in workplace power and economic constraints. Women are more likely, on average, to carry substantial caregiving responsibilities outside the workplace, so a health problem or an unpredictable schedule may make continued employment more difficult to sustain. Reducing hours or exiting work may be an individual response to strain, but it can also carry long-term costs, including lost wages, weaker access to employer-sponsored benefits, reduced retirement contributions and diminished career progression. Men who remain employed despite deteriorating health may face a different risk: their continued presence in the workforce can conceal the severity of the burden until illness becomes more advanced. The study therefore presents schedule instability as a condition that can produce unequal forms of disadvantage rather than a uniform experience shared by all workers.
The researchers say the results broaden the meaning of job quality. Pay, benefits and total hours remain important, but the predictability of those hours may also influence health and employment stability. Policies and workplace practices that provide advance notice of schedules, limit extreme month-to-month fluctuations or give workers greater control over their hours could reduce some of the strain associated with volatile employment. The study’s broader warning is methodological as well as social: when people reduce their hours or leave work because their health is failing, those changes should not be treated as statistical attrition. They are part of the outcome. By following workers beyond continuous employment, the analysis exposes a wider and more consequential picture of how precarious schedules affect health—and why the workers most harmed may be the least visible in traditional employment research.
Subject of Research: People
Article Title: Unstable hours, unequal health: Gendered occupational health costs of work-hour volatility in the U.S.
Web References: https://sociology.illinois.edu/ ; https://sociology.illinois.edu/directory/profile/tfliao ; https://doi.org/10.1016/j.ssmph.2026.101949
References: Lu, Sinyee Qianyi, and Tim Liao. “Unstable hours, unequal health: Gendered occupational health costs of work-hour volatility in the U.S.” SSM – Population Health. DOI: 10.1016/j.ssmph.2026.101949
Image Credits: Photo by Michelle Hassel
Keywords: work-hour volatility, occupational health, unstable schedules, precarious employment, gender differences, self-rated health, employment instability, workplace health, labor force participation, public health
